Electronic repositories have transformed how library and information management professionals in Kenya access, manage, and disseminate knowledge resources. Effective information retrieval from these digital collections is essential for enabling users, whether researchers at universities or staff at county government offices, to locate relevant documents swiftly and accurately. This chapter explores practical techniques for performing information retrieval within electronic repositories, including how to manage and export search results in various formats suited to different professional needs.

3.1 Definition of term

Information retrieval in an electronic repository is the process of locating, from an organised collection of digital resources, the items that answer a user's information need. An electronic repository is a managed digital store, such as an institutional repository, a digital library or a library database, in which theses, journal articles, reports, e-books and other digital items are stored with descriptive data (metadata) so that they can be found, accessed and preserved. A library and information professional must be able to define the terms used in this work precisely, because users, colleagues and system documentation all rely on the same vocabulary. Example: when a student at a college library asks for "past research on cataloguing of digital theses", the librarian first understands what is meant by repository, query and search strategy before choosing how to search.

Search Strategy

A search strategy is a planned, step-by-step approach for finding information in one or more electronic sources. It states what the user wants to find, which words and concepts will represent that need, where the search will be run, how the search statements will be built, and how the results will be checked and improved. A good strategy saves time and reduces missed or irrelevant items, because the search is designed before it is typed rather than guessed at in the search box.

Example: a postgraduate student needs material on how Kenyan college libraries organise digital theses. Instead of typing one long sentence into the repository, the librarian writes down the main concepts (digital theses, cataloguing, college libraries), lists alternative words for each, decides to search the institutional repository first and a specialised database second, runs the search, reviews the first results and then adjusts the words used.

The main elements of a search strategy are as follows:

• Information need: A clear statement of what the user is looking for, why it is needed and how much detail is required. Without it, the search has no direction and results cannot be judged.
• Main concepts: The key ideas in the need, usually two to four, broken out of the full question. For the cataloguing example, the concepts are digital theses and cataloguing.
• Search terms: The keywords, synonyms, related terms and spelling variants that express each concept, for example "thesis", "dissertation" and "electronic theses".
• Choice of sources: The repository, catalogue or database in which the search will be run, selected according to the subject and type of material needed.
• Search statements: The way the terms are combined and entered, together with any limits such as language, date or material type.
• Evaluation of results: A check of how many items were found and how relevant they are, to decide whether the search is too broad, too narrow or acceptable.
• Refinement and record keeping: Changing terms or limits where results are poor, and noting the final strategy so that the search can be repeated or shared with the user.

Related terms used in information retrieval

Several other terms appear alongside search strategy and are used throughout this chapter. Knowing them helps a librarian explain results to users and write clear search records.

• Query: The words or statement that a user enters into a search box to describe what is wanted. A query is the written form of the information need.
• Metadata: Descriptive data about a digital item, such as title, author, subject, date and abstract. Searches mostly work by matching the query to metadata and, where available, to the full text.
• Index: An organised list of terms taken from the items in a repository, linked to the records that contain them. The system searches the index rather than every document, which makes retrieval fast.
• Relevance: The degree to which a retrieved item actually answers the user's need. An item can match the words typed but still be irrelevant to the purpose.
• Precision: How many of the items retrieved are relevant. A search with high precision returns few useless results.
• Recall: How many of the relevant items in the repository were successfully retrieved. A search with high recall misses very little.
• Retrieval set: The list of records returned by a search, which the librarian reviews, sorts and, where needed, narrows down.

Chapter Summary

This chapter focused on performing effective information retrieval within electronic repositories by first defining user needs to ensure search relevance and accuracy. It emphasized selecting appropriate search tools such as search engines like Google Scholar and PubMed, library catalogues including OPACs, and specialized databases like JSTOR and IEEE Xplore to match the information requirements. Various search techniques were covered, including keyword and phrase searching, the use of Boolean operators such as AND, OR, and NOT, and advanced options like field-specific searches and date range filters to refine results. The chapter also addressed the importance of exporting search results in multiple formats including PDF, CSV, and RIS to facilitate further use and management. Additionally, it highlighted the role of gathering electronic information feedback from users to assess the effectiveness of search strategies. Finally, the chapter described how to implement improvements based on this feedback by adjusting search methods, updating training materials, and conducting user training to enhance overall retrieval performance.